A primer on bank reserves

The $1 trillion expansion of the Federal Reserve's Balance Sheet that occurred from Sept. '08 through late last year is arguably one of the biggest monetary events to happen in the history of the U.S. This chart puts it into perspective (note that the y-axis is a logarithmic scale). Prior to this crisis, bank reserves totaled about $96 billion and hadn't grown at all since 2003; now they stand at $1,190 billion.

I like to keep big and complex things as simple as possible in order to better understand their significance, so I offer this simple description of what has happened and what it might mean. It's not meant to be a definitive analysis of the situation, but rather something that should be understandable to those with a minimal level of knowledge of monetary matters. And at the very least it provides a basis for discussion.

The Fed, responding to a huge increase in the world's demand for money and safety in the wake of the collapse of Lehman Bros., bought a trillion dollars worth of securities (mostly Treasuries and MBS, and the exact number is closer to $1.1 trillion). The Fed paid for these purchases by crediting the accounts of its member banks with reserves, which is the type of money only the Fed can create. Buying and selling securities is the traditional way that the Fed increases or reduces the supply of money to the banking system.

In short, the Fed reacted to an explosion in the demand for money by pumping an explosion of bank reserves into the system. This was the right thing to do, since to not accommodate a surge in the demand for money with a surge in the supply of money would almost certainly have resulted in a severe shortage of money and thus a monetary deflation of the sort that exacerbated the Great Depression.

To date, banks essentially haven't used any of their extra reserves to expand their lending. The reserves are sitting idle at the Fed in the form of "excess reserves," which currently total $1.06 trillion, up from roughly zero prior to Sep. '08. The Fed has increased the lending capacity of banks by an order of magnitude, but this hasn't created a flood of extra liquidity or a burst of inflation.

There are several reasons for the lack of lending. To begin with, the demand for money and safety remains high, and thus the demand for loans is still weak. To be sure, there are lots of companies that are desperate for credit to fund startups and expansions that can't find a bank willing to lend to them, but in aggregate, bank lending is weak in part because most people these days are still trying to deleverage. At the same time, banks aren't particularly anxious to lend either. They still are suspicious of the credit quality of borrowers, which is why they have raised their lending standards, and they are more risk-averse than usual, just like almost everyone. Banks aren't anxious to change this, being quite happy to keep overall risk low even if it means earning a piddling amount on all the reserves they have at the Fed.

Banks may someday decide to use their reserves, and if and when they do, they could make tons of new loans and print tons of money in the process. (This is how the fractional reserve banking system works: if you ask the bank for a loan of $1 million, the bank sets aside about $100,000 in reserves and credits your checking with $1 million.) If this were to happen it could be very inflationary. For example: $1 trillion of excess reserves could potentially support about 10 trillion of new bank deposits--a sixfold increase in bank deposits!

The Fed is naturally quite concerned about this possibility, and the markets are too. Ideally, the Fed would just reverse what they did in late 2008, and sell Treasuries and MBS in exchange for taking back the trillion dollars of excess reserves. But everyone worries that this could push interest rates sky-high, threaten the recovery, and take away the security blanket (i.e., the tons of excess reserves) that now keeps the banks warm and happy. Alternatively, as Bernanke argued in a WSJ article last summer, the Fed could allow the reserves to gradually decline over the years as the securities it holds mature or are prepaid, but this might leave too many reserves in the system for too long.

Bernanke now says that the Fed's preferred exit strategy is to continue to pay interest on reserves (an authority first granted to the Fed in the Fall of '08), even as the Fed raises short-term interest rates. That way, the theory goes, banks won't mind holding lots of excess reserves indefinitely, or at least holding a lot more than they would under a non-interest-paying regime. Bank reserves would thus become very much like T-bills, a source of bedrock security and income for banks. As the Fed ratcheted up short-term interest rates to keep inflation pressures from rising as the economy gains strength, the income on these reserves would grow commensurately. Banks would be happy to keep holding excess reserves, and the Fed wouldn't have to sell a trillion dollars of securities. The Fed would presumably pay an interest rate on reserves equal to a bit less than the desired Fed funds target rate, which in turn is what drives interest rates all along the Treasury yield curve. Between maturing debt and mortgage prepayments, the excess reserves would gradually disappear.

Without the ability to earn interest on their reserves, reserves would act like a deadweight on bank's balance sheets. Banks would either try harder to expand their lending, which could be very inflationary, or they would sell their excess reserves on the open market. The latter would severely depress the Fed funds rate, which is something the Fed doesn't want to happen. The Fed runs monetary policy by targeting the Federal funds rate, and they are going to want to raise that rate, not watch it fall.

Alternatively, the Fed could raise reserve requirements, since that would effectively soak up some portion of the excess reserves. But raising reserve requirements can't do the heavy lifting that paying interest on reserves can accomplish, because being forced to hold a lot more non-interest-paying reserves would again be a deadweight drag on banks' balance sheets.

So Bernanke's proposal to pay interest on reserves is not completely crazy. But it could become very expensive for the Fed if interest rates rose a lot, and the whole exercise takes us into uncharted waters where unintended and unforeseen consequences lurk in the depths. It would also act to legitimize the "monetization of government debt" that is the real source of all inflation.

We can only hope that the Fed is able to navigate the turbulent and uncharted waters that lie ahead, while avoiding Scylla (inflation) and Charybdis (deflation).